International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
RELEASE OF ASH-FORMING ELEMENTS FROM BIOMASS
DURING OXIDATION AND PYROLYSIS
Anders Brink*A, Johan WerkelinA, Liang WangB, Ehsan HoushfarC, Terese LøvåsC
and Mikko HupaA
A) Åbo Akademi University, Finland
B) SINTEF Energy Research, Norway
C) Norwegian University of Science and Technology, Norway
* Corresponding author: anders.brink@abo.fi
ABSTRACT
In this work, the release of several ash-forming elements from four biomass fuels has
been quantified as function of temperature and residence time in a single particle
reactor. The fuels are spruce bark, torrefied softwood, wheat straw and miscanthus. For
the experiments, the fuels were pelletized. In the experiments, three different
temperatures were used 800, 900, and 1050 °C. In addition, two different gas
atmospheres were used, one containing 3% O2 and one oxygen free. The release of the
ash forming elements was determined as a function of conversion. For the pyrolysis
experiments the same holding times were used in order to see the influence of the gas
atmosphere. The release was calculated using the elemental composition of the chars
analyzed using ICP-MS and ICP-AES.
Keywords: Release; biomass conversion
1
Introduction
Understanding the behaviour of ash-forming elements is the key to predict operational
problems such as agglomeration, slagging, fouling, particulate formation and corrosion
related problems. Many of these aspects are discussed in the review paper of Werther et
al. [1]. Understanding the release of the ash-forming elements can also be valuable
when it comes to operating the combustion facility in such a way that the ash can be
further utilized, for example in cement manufacturing or as a fertilizer.
The release of ash-forming elements during thermal conversion of biomass has been
investigated in a number of studies. Most attention has been put on the K, S and Cl. A
number of techniques have been applied, the probably most common one though is a
fixed bed reactors setup [2-4], but also flow reactor [5,6] and grid heater setup has been
utilized [7,8].
In the present study the release of ash forming elements during thermal conversion of
single pellets are studied. The release is studied both at combustion as well as at
pyrolysis conditions. The aim of this study is to reveal the influence of temperature,
atmosphere composition and char conversion on the release of a number of ash forming
elements.
-1-
International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
2
Experimental
2.1 Laboratory experiments
Pellets were prepared from pulverized/grinded fuel. Four different fuels were studied:
spruce bark, torrefied softwood, wheat straw and miscanthus. The pellets had a diameter
of 8 mm. The height of the pellets varied, since the controlling parameters were the
amount of fuel, in this case 200 mg, and the applied pressure.
The pellets were placed on a sample holder of quartz glass. Before being inserted into
the hot gas atmosphere the samples were kept in a protective N2 atmosphere. Then the
samples were quickly inserted. After a certain holding time the sample was removed
from the hot environment into the cold protected N2 atmosphere. Here the sample was
allowed to cool down to room temperature before being removed from the quartz
holder.
The test matrix included at total of 96 experiments: four fuels, two gas atmospheres,
three temperatures and four different holding times. The gas atmospheres were 1) 3%
O2 + 97% N2 and 2) 100% N2. The temperatures were 800 °C, 900 °C and 1050 °C. The
holding times were 1) end of pyrolysis, 2) 50% char burnout time, 3) char burnout and
4) burnout + 5 min cooking time. The experiments were performed such that first the
longest holding times were studied. During these runs the devolatilization time was
established based on flame out, the burn out time was established using CO and CO2
analyzers. The 50% burnout time is taken as the mid-point between end of pyrolysis and
end of char oxidation. For the experiments performed in N2 the holding times
established in the runs with O2 in the atmosphere were utilized.
2.2 Fuel and char analysis
The inorganic content of fuel, and char and ash samples were analyzed by two different
methods: ICP-AES and ICP-MS. ICP-AES stands for inductively coupled plasma
atomic emission spectroscopy. The analyzer consists of two parts: the plasma
generating part and the optical unit where the intensity of the lines in the atomic
emission spectra is analyzed. The ICP-MS differs from the ICP-AES in that a mass
spectrometer is used for quantification. The detection limit of the ICP-MS is lower than
that of the ICP-AES. On the other hand, the ICP-AES is a robust method for analyzing
non-metals like Si, S and P.
For digestion of the solid samples, micro-wave assisted digestion was utilized. The
samples they were first digested in a mixture of HNO3 (4ml), HClO4 (2ml) and HF
(0.5ml). For some of the samples also 2ml of H2O2 were added. The amount of digested
sample varied from approximately 200 mg for the untreated fuel samples to less than 1
mg from the torrefied fuel char residue. In order to digest the samples fully they were
then kept 60 minutes in the acids at elevated temperatures.
-2-
International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
3
Results
The results from the elemental analysis were recalculated to mass element/mass of
untreated fuel. The analysis results obtained with the two different methods agreed well.
Figure 1 shows results for two elements, but similar trends were observed for all
elements both methods could analyze. Since the ICP-MS could analyze a larger number
of elements, results obtained using this method will be focused on in this paper.
1000
1200
Mg
900
Al
1000
700
ICP-AES μg/g
ICP-AES μg/g
800
600
500
400
300
800
600
400
200
200
100
0
0
200
400
600
800
1000
0
0
200
400
ICP-MS μg/g
600
800
1000
1200
ICP-MS μg/g
Figure 1. Element analysis results from ICP-AES vs. ICP-MS.
The results from the analysis showed that most elements were retained in the char
residue/ash regardless of the conditions. This was the result for the following elements:
Ca, Fe, Mn, Cr, CO, Ni, Cu, Na, Mg, Al, Pb, Li, Rb, Sr, Cd, Ba and As. Many of these
elements were present in concentrations less than 100 μg/g untreated fuel: Cr, Co, Ni,
Cu, Pb, Li, Rb, Sr, Ca and As. For some of the elements the analysis showed
considerable variations, e.g. for Fe and Na the variations were considerable.
The most interesting trends were observed for K and Zn. Figure 2 shows the potassium
content on fuel bases for the four fuels at different conversion stages. The figure shows
that for bark and torrefied wood there is a release of K at the highest temperature for the
case of complete char burnout plus 5 min cooking time. This is partly in contrast to the
findings of Johansen et al. [3] that found that the release of K toke place primarily in
connection to the sublimation of KCl at temperatures exceeding around 700 °C. The
differences in the results can most likely be explained by differences in the experimental
setup and conditions. For the herbal fuels no release of K seems to occur. These fuels
clearly also has the highest potassium content. It is interesting to note that the straw has
clearly the highest content of Si, approximately 12000 μg/g untreated fuel, whereas the
content of Si in the torrefied wood is only around 300 μg/g untreated fuel. The bark and
the miscanhtus both have similar content of Si, around 4000 μg/g untreated fuel, but for
the bark the scatter is rather large. This may indicate that released potassium is trapped
in a silicate phase in the ash of the straw and miscanthus as discussed by others [3].
-3-
International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
2000
Bark
Torrefied wood
900 C
1600
Fuel
Dev
3% O2
N2
1050 C
1600
1400
1400
1200
1200
1000
800
600
600
400
400
200
200
0
1
2
3
4
5
6
7
8
10000
2
N2
3
4
5
6
7
1050 C
8
Fuel
Dev
3% O2
N2
1050 C
8000
7000
6000
6000
5000
4000
3000
3000
2000
2000
1000
1000
2
3
4
5
6
7
0
8
900 C
Fuel
Dev
1
2
3% O2
N2
1050 C
5000
4000
1
800 C
9000
900 C
g/g
g/g
1
3% O2
Miscantus
800 C
7000
0
Dev
10000
Straw
9000
8000
900 C
Fuel
1000
800
0
800 C
1800
g/g
g/g
2000
800 C
1800
3
4
5
6
7
8
Figure 2. Potassium content of the four fuels as μg/g untreated fuel. In each plot,
Group “1” are results for the untreated samples, “2”are the devolatilized samples,
“3” and “6”the 50% burn out, “4”and “7” the 100% burn out and “5” and “8” 100%
burnout + 5 min cooking time. In the figure the atmosphere is also indicated. The red
line indicates the average content in the untreated fuel samples.
Figure 3 shows the zinc content on fuel bases for the four fuels at different conversion
stages. Here the trend for the bark and the torrefied wood is very clear: zinc is
effectively released at reducing conditions. At oxidizing conditions the zinc release
increases with temperature. The results for the devolatilization tests are surprising and
require further investigation. One possible explanation for the results is that zinc is
released after the sample has been retracted into the protective N2 atmosphere. For
miscanthus and straw the trend is difficult to interpret because of scatter in the analysis.
All but one analyzed miscanthus sample has low zinc content. Also the zinc content of
the straw samples are generally low according to the analysis results, expect for two
samples prepared at 800 °C and in an oxidizing atmosphere.
-4-
International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
100
100
Bark
Torrefied wood
90
Fuel
Dev
3% O2
N2
70
1050 C
50
30
20
20
10
10
3
4
5
6
7
0
8
100
1
2
3
4
5
6
7
8
Fuel
Dev
3% O2
N2
1050 C
80
70
60
60
50
40
30
30
20
20
10
10
2
3
4
5
6
7
0
8
Dev
1
2
900 C
3% O2
N2
1050 C
50
40
1
Fuel
800 C
195
90
900 C
g/g
g/g
1050 C
Miscanthus
800 C
70
0
800 C
100
Straw
90
80
N2
50
40
2
3% O2
900 C
30
1
Dev
60
40
0
Fuel
70
900 C
60
g/g
80
800 C
g/g
80
90
3
4
5
6
7
8
Figure 3. Zinc content of the four fuels as μg/g untreated fuel. In each plot,
Group “1” are results for the untreated samples, “2”are the devolatilezed samples,
“3” and “6” the 50% burn out, “4”and “7” the 100% burn out and “5” and “9”
100% burnout + 5 min cooking time. In the figure the atmosphere is also indicated.
4
Conclusions
The fate of ash-forming elements in four biomass fuels during conversion has been
studied in two different atmospheres in the temperature range 800 °C to 1050 °C. The
results shows that most of the metallic elements only present in trace amount are
retained in the fuel sample. Also some of the elements present in slightly higher
amounts, such as Ca and Mg, are also retained in the fuels. Potassium on the other hand
is released from the bark and the torrefied wood. The release increases with temperature
and holding time. This behavior is not observed for the miscanthus and the straw
samples. Zinc is released from the bark and the torrefied wood at reducing conditions.
At oxidizing conditions zinc release could also be observed at high temperatures.
5
Acknowledgements
The experiments in this study have been carried out with support from the BRISK
project funded by the EU FP7 program. Additional funding from the Academy of
Finland project Symbiosis is also acknowledged.
-5-
International Flame Research Foundation
The Finnish and Swedish National Committees
Finnish – Swedish Flame Days 2013
REFERENCES
1. Werther, J., Saenger, M., Hartge, E.-U., Ogada, T., Siagi, Z., Combustion of
agricultural residues, Progress in Energy and Combustion Science 26 (2000) 1–27.
2. Knudsen, J.N., Jensen, P.A., Kim Dam-Johansen K., Transformation and Release to
the Gas Phase of Cl, K, and S during Combustion of Annual Biomass, Energy &
Fuels 2004, 18, 1385-1399.
3. Johansen, J.M., Jakobsen, J.G., Frandsen, F.J., Glarborg, P., Release of K, Cl, and S
during Pyrolysis and Combustion of High-Chlorine Biomass, Energy Fuels 2011,
25, 4961–4971.
4. Davidsson, K.O., Korsgren, J.G., Pettersson, J.B.C., Jäglid, U., The effect of fuel
washing techniques on alkali release from biomass, Fuel 81 (2002), 137-142.
5. Sippula O., Lind, T., Jokiniemi, J., Effects of chlorine and sulphur on particle
formation in wood combustion performed in a laboratory scale reactor, Fuel 87
(2008) 2425–2436.
6. Shah, K.V., Cieplik, M.K., Betrand, C.I., van de Kamp, W.L., Vuthaluru, H.B.,
Correlating the effects of ash elements and their association in the fuel matrix with
the ash release during pulverized fuel combustion, Fuel Processing Technology 91
(2010) 531–545.
7. Kymäläinen, M., Janka, K., Keskinen, J., Moisio, M., Backman, R., Hupa, M.,
Measurement of time-dependent fume release rate during black liquor pyrolysis,
Journal of Pul and Paper Science 22 (1996), 17-23.
8. Okuno,T., Sonoyama,N., Hayashi, J., Li, C.-H., Sathe, C., Chiba, T., Primary
Release of Alkali and Alkaline Earth Metallic Species during the Pyrolysis of
Pulverized Biomass, Energy & Fuels 2005, 19, 2164-2171.
-6-